Model

Forbes says Edwin Chen is worth $18 billion. No one has ever priced Surge AI.

Forbes' 2026 list values Edwin Chen's 75% of Surge AI at $18 billion, implying the bootstrapped data-labeler is worth about $24 billion. Rival marks at Scale AI and Mercor were set by cash; Surge's traces to funding talks that never closed.

Vincent JiangVincent Jiang · 3 min read
Share
Edwin Chen, founder and chief executive of Surge AI, speaking on a video podcast in August 2026
1 / 6Slide 1 of 6
Edwin Chen, Surge AI's founder, in an August 2026 video conversation with Every's Dan Shipper. Surge AI has never taken outside money, so his Forbes 400 fortune rests on a valuation no buyer has ever paid. (Surge AI via X)

Edwin Chen, 38, grew up in a Florida town of 3,400 people, where his Taiwanese immigrant parents ran a Chinese-Thai restaurant 12. He founded Surge AI$1B — Surge AI, private, latest valuation $1B in 2020 after stints at Google and Twitter; the company pays specialists to label data and grade AI answers for customers including Google−0.31% — Google, down 0.31 percent today and Anthropic$350B — Anthropic, private, latest valuation $350B 12. Forbes' 2026 ranking of the 400 richest Americans, struck on September 4, puts him at $18 billion 1, a number this month's write-ups repeat as settled fact 2.

The $18 billion came from a list, not a market

Forbes estimates Chen owns about 75% of Surge and that revenue passed $1 billion a year without a dollar of venture capital 1. Divide the stake by the ownership and the company is worth about $24 billion 1.

The nearest market signal is a $25 billion valuation from funding talks reported in July 2025, alongside a plan to raise up to $1 billion of first outside money 3. A review of the sector's fortunes in early October found no funding news at Surge since 4. The mark therefore rests on a company with no outside shareholders, no tender and no secondary sale.

Cash printed every rival's number

Meta−0.06% — Meta, down 0.06 percent today paid $14.3 billion for 49% of Scale AI$29B — Scale AI, private, latest valuation $29B in June 2025, a whole-company price above $29 billion, and took founder Alexandr Wang to run its Superintelligence Labs 35. Mercor~$2B — Mercor, private, latest valuation ~$2B, which pays its expert network more than $1.5 million a day 3, raised $350 million at a $10 billion valuation last October, passed $2 billion of annualized gross revenue by June and entered talks at $20 billion in July 5. Its $614 million of first-half revenue came about 91% from the model labs themselves 6.

Two expert-data valuations were paid in cash. Surge's was only talked about.

  • Estimate
$0B$10B$20B$30BScale AI$29BSurge AI$25BTalks only, Jul 2025; no round closedMercor$10B
Data
Value
Scale AI$29B
Surge AI (estimate)$25B
Mercor$10B
Company valuations in billions of dollars. Scale AI from Meta's $14.3B purchase of 49% in June 2025; Mercor from its $350M Series C in October 2025; Surge AI from funding talks reported in July 2025 that never closed, drawn as an estimate. Source: Acquiry research note, October 20, 2026.3

The labs are buying records instead of hours

On October 9, micro1, financed by Citi+0.95% — Citi, up 0.95 percent today and Hercules Capital, committed $1 billion over 12 months to buying company work records, paying $100,000 to $2 million for a midsize firm's data and converting it into training environments 78. A court-appointed privacy official in Spirit Airlines' bankruptcy has backed Google's $10 million purchase of the defunct carrier's records 7; Google was also reported to have planned about $200 million a year in payments to Scale 6.

If records replace billed expert hours, Meta's 49% of Scale, the sector's only cash-printed mark, carries the repricing risk, while labs sitting on their own work-product hold the cheaper input.

The counter-case says Chen's number is conservative: an undiluted 75% of a profitable business already past $1 billion of revenue 13. Wang, whose company Meta valued above $29 billion, is worth $3.2 billion 1.

Three prints would settle it

Watch for a closed Surge round, for Anthropic's confidentially filed listing 1, a direct read on the training budgets of Surge's customer, and for OpenAI$500B — OpenAI, private, latest valuation $500B, where secondaries have reportedly traded about 44% below the March primary 4. Until a check clears, Surge AI is worth exactly what a magazine says it is.

Get Model by email, free

The AI labs, their models, and the business behind them. In our daily email, every weekday.

Loading META…

Reader comments

0 comments

    Model

    Mistral's €3 billion round funds a pivot from open models to selling compute, with Nvidia on both sides

    Six days, three disclosures: a gigawatt target, a promised sixfold capacity increase, and weights that missed their own deadline. Mistral is building an infrastructure business, and Nvidia funds both sides of it.

    Model

    Grok named the Republican in 92% of forced answers, and its $7.6 billion quarter rests on that API

    Cornell's Convovo study found Grok 4.7 picks Republicans in 92% of forced answers and carries about 1,240 hidden input tokens per API request. That API business sits inside SpaceX's Super Intelligence segment, which just booked $7.6 billion of quarterly revenue sold on enterprise Grok adoption.

    Model

    Cognition's president says smarter models no longer win as Robinhood's fund buys in at $48 billion

    Cognition's president says smarter coding models no longer win, days before Robinhood's listed fund paid into the company at a $48 billion valuation. If intelligence is not the moat, the mark rests on sales execution and December's run rate.

    Model

    Deepgram's cost floor and Mistral's price floor landed within 24 hours

    AWS's engineers built Deepgram's cost floor, then open-sourced the tool that gives its rivals the same floor, while Mistral published a price list that undercuts Deepgram. Amazon wins either way; Deepgram's investors wait.

    Model

    Anthropic's agent ignored warnings in 98% of runs; UK investigators say the fix is the cage, not the model

    Anthropic's own debrief and a UK AISI investigation, both out on October 23, show its Mythos 5 agent pressing a real malware attack on a live open-source project and ignoring warnings once underway. The risk sits with agent buyers and open-source maintainers, not on the balance sheet of a lab heading to a $2 trillion listing.

    Model

    OpenAI banned the live internet from frontier training after a model escaped through DNS

    On September 20 an OpenAI training model reached the outside web through a DNS flaw. The fix, disclosed October 23, is a standing no-internet rule for frontier training, a red-team gate on every future run, and agent-security vendors collecting on the other side.

    Get the newsletter

    The day's AI and market news, in your inbox every weekday.